Just a post for anyone who is interested in the differences between dommed and dedommed/delensed on the SBT90.2 in the Convoy L8.
There was about a 30% difference in hot spot size with it being smaller and more intense after dedoming but also a noticeable increase in brightness.
ummmm Luminus SBT90 doesn’t have a dome? Maybe It’s Cree XHP70.x HD you’re talking about
It has a flat glass lens and a metal frame to support it, which both lead to an increase of the effective LES size by causing diffusion.
This is wild! I’ve always known that there is an improvement, but a hotspot size difference by a factor of 1.3 (thus an intensity difference of 1.69) is insane, and I don’t think fully explained by the difference in apparent LES.
I pulled up images of an original and delensed sample from djozz’s test thread, and running a profile plot found less than 1% of difference in the side length of the yellow region.
The glass window is certainly diffusive, but I don’t think by a factor of 1.3. Maybe the delensing also shifted (and inadvertently optimized) the focus?
It would have been nice to get some quantitative intensity measurements, but sadly it’s too late for that now…
Koef3 mentioned that light gets reflected by the metal frame, thus creating a new “origin” for a part of the light that’s pretty far off the hotspot. I don’t think it was that much of a difference either, however.
Edit: found it:
I see it in the under-exposed photo of the emitter! The frame reflection can partially explain why the de-lensed beam looks cleaner.
6-8% luminance is nowhere near 30% change in hotspot size–I’m inclined to chalk that up to imprecise photography: it’s clear that the two beamshots are taken from slightly different locations/distances, which could prompt the camera to overexpose one more than the other.
WB, PaulorNuthn!
When the show The Wonder Years was popular, I looked like the character Paul.
When the movie Napoleon Dynamite was popular, I looked like the title character.
Now I look like John Oliver from Last Week Tonight.
I’m a nerd that has always looked like a nerd. 
Hi, thanks for sharing. I’m surprised at the huge difference in hotspot size. Was the 30% based on the pictures posted? Asking because it seems the 2 pictures are not to scale. Left pic is magnified; for example width of wood panel on left is larger than right.
Another thing I noticed is the beam of right side is at an angle. Were the flashlights equidistant from the wall?
The pics were taken a day apart, just sitting on my bed, phone in one hand and flashlight in the other.
The 30% is an estimate from a real time view, the hotspots actually look half the size as the photos show, must be from phone settings or exposure.
I’m no camera expert and this certainly wasn’t an official scientific experiment under strict conditions.
Just posting what I captured and my experience of it in case it’s helpful to anyone wondering.
I don’t work for Convoy or any LED manufacturers.
Edit: Yes it’s a glass frame, just get a blade under it and approach from the - cathode side as its unglued there, once the blade is under the corner just twist and it pops off.
It definitely ended up brighter, smaller more intense hot spot, centred the beam a bit more and reduced the carona by about half.
All not clearly seen in photos unfortunately.
Doesn’t Hank de-lens his sbt 90’s?
Making me even more tempted to de-lens my sbt90.2 and nbt160 for that smaller hotspot!
Sadly the only good host I have for them is the FFL T9R and it uses the lens to center the reflector.
Same as the L8, but seeing as it already was off centre from manufacturer I figured it couldn’t get much worse.
So after delensing, I was careful to make sure the reflector screwed back on without touching any of the bond wires.
I did that by holding the flashlight upside down and over my head as i looked straight up into it.
Then I slowly started screwing the bezel back on and tapping it left or right to keep centering the reflector until it was tight enough to push against the star and restrict further movement.
It’s not perfect but much better than it was.
Ah, I see.
Yeah I guess I could try installing it that way too, but not gonna try it till I have spare sbt90.2 haha
Not a bad idea.
